In present agricultural practices, the lack of water and effective resource management systems has emerged as major obstacles. A potential solution to these problems can be achieved by combining smart irrigation systems and intelligent decision-making. This study investigates how machine learning algorithms can be used to improve conventional irrigation systems. The proposed smart irrigation system dynamically adjusts irrigation schedules to ensure adequate watering by leveraging real-time data from soil moisture sensors, humidity sensors, temperature sensors, and crop-specific attributes. Results of the proposed smart irrigation system have shown verifiable benefits over traditional counterparts by significantly decreasing water usage and improving overall crop yield.